Journal of Hydrologic Engineering

Radar and multisensor quantitative precipitation estimates (QPEs) have played a critical role in real-time hydrologic and weather predictions. The utility of this data set for hydrologic model calibration, however, is hampered by the limited data archive and the presence of data gaps. In the research reported in this paper, the use of a composite reflectivity data set is investigated, namely the Weather Services International (WSI) National Operational Weather radar (NOWrad) data set, to create an hourly QPE that would complement the National Weather Service (NWS) Stage III archive by filling in the latter’s gaps and potentially improving the latter’s accuracy. The paper starts from an inventory analysis of Stage III and NOWrad products, and through which it is found that as much as 17% of the Stage III data were missing for some of the regions, and some of these gaps can be filled using the NOWrad data. Then, the paper presents two experimental products, as follows: (1) NOWrad–Raw, based on variable $Z(\mathrm{reflectivity})-R(\mathrm{rain}\mathrm{rate})$ relationship derived for each month using monthly Cooperative Observer Porgram (COOP) station totals; and (2) NOWrad–bias-corrected mosaic (BMO), derived by bias-correcting the NOWrad–Raw using hourly rain gauge products. Then the two products along with Stage III products against independent hourly gauge reports over two locations [(1) Charlotte–Mecklenburg metropolitan area in North Carolina, and (2) Lower Colorado River drainage in central Texas]. The analyses over the two test sites reveal the following: (1) NOWrad–Raw product suffers from a negative overall and conditional bias, while the Stage III is closer to bias-neutral; and (2) NOWrad–Raw product, at least over the Charlotte–Mecklenburg area, can outperform the Stage III in terms of correlation with gauge data and skill in detecting light rainfall. Further analyses of the Raw and bias-corrected NOWrad suggest that negative conditional bias may be substantially improved, but there is a tendency to overcorrect the bias for the summer months, possibly due to the presence of false detections. The results show that NOWrad can be a viable source of high-resolution quantitative precipitation information and it complements the current NWS archive in several respects. Possible mechanisms for further improving the accuracy of the NOWrad QPE are discussed.

Assessment of radar-based precipitation estimates using rain gauge observations is a critical exercise in evaluating pre-and postcorrected (gauge-adjusted) radar-based precipitation data. A comprehensive assessment framework combining several visual, quantitative, and statistical measures, indexes, and skill scores is proposed and developed for evaluation of radar-based precipitation estimates in space and time. Contingency measures, skill scores, and a few new metrics are proposed and are evaluated along with several indexes. Visual measures provide a quick check of agreement between radar and rain gauge data sets. Quantitative measures provide information about errors, and skill scores assess the quality of radar data for dichotomous (rain and no-rain) events. Summary statistics and hypothesis tests in statistical categories provide insights into distributional aspects of the rain gauge and radar data sets. The framework is used for evaluation of 15-min radar-based precipitation data obtained from the South Florida Water Management District (SFWMD). Four years of radar and rain gauge data available at 189 sites are used for analysis. Results suggest that radar data in the SFWMD region have progressively improved during the period of analysis. All indexes and skill scores used in the current study suggest that radar data are of good quality at different temporal resolutions and in agreement with rain gauge data. However, spatial bias evaluation suggests that radar data underestimate precipitation amounts in two areas of the SFWMD region.

In recent years, there has been a growing interest by different user communities to use radar-based rainfall products for identifying and quantifying heavy precipitation and other related information. A potentially viable product for such analyses is the National Weather Service (NWS) Stage IV Quantitative Precipitation Estimate (QPE), which covers the entire conterminous United States and is available at 1, 6, and 24 h temporal resolutions. The current article focuses on data artifacts in the Stage IV product over its full record period (2002–2013) and assesses their impacts on the analysis and derivation of heavy rainfall statistics. Examining annual maximum series extracted from the Stage IV 1-h accumulation product revealed the presence of numerous artifacts that were mostly linked to erroneous rain gauge reports and which can be mistaken as rainfall maxima. Artifacts were found over many parts of the United States, including the west, south, southeast, and northeast. Other erroneous features that affect the spatial delineation of heavy precipitation include nonrainy lines and polygons in many regions over the product domain. The current analysis demonstrated that these artifacts had negative impacts on the estimation of parameters and quantiles of commonly used extreme value probability distributions. Additional exploratory analyses indicate that Stage IV 6-h and 24-h accumulations, as well as the longer-latency QPE archive from the regional River Forecasting Centers (RFCs), are mostly free from the artifacts shown in the Stage IV 1-h product. These comparative analyses revealed that the Stage IV product, at least over some of the RFC domains, misses on the benefits of human quality control that is introduced at a later stage at the RFCs, and suggested that caution be exercised in applying Stage IV as the basis for analysis of heavy precipitation.

Weather radar data used for urban drainage applications are traditionally adjusted to point ground references, e.g., rain gauges. However, the available rain gauge density for the adjustment is often low, which may lead to significant representativeness errors. Yet, in many urban catchments, rainfall is often measured indirectly through runoff sensors. This paper presents a method for weather radar adjustment on the basis of runoff observations ($Z\text{-}Q$ adjustment) as an alternative to the traditional $Z\text{-}R$ adjustment on the basis of rain gauges. Data from a new monitoring station in Aalborg, Denmark, were used to evaluate the flow-based weather radar adjustment method against the traditional rain-gauge adjustment. The evaluation was performed by comparing radar-modeled runoff to observed runoff. The methodology was both tested on an events basis and multiple events combined. The results indicate that this $Z\text{-}Q$ adjustment method performs similarly to $Z\text{-}R$ adjustment on the basis of well-placed rain gauges. This opens up the possibility of using flow measurements as an alternative to or in combination with rain gauges.

Multidecadal records of hourly precipitation estimates are needed to provide forcings for the simulation of subdaily processes in hydrologic models. Existing climatological datasets, such as the North American Land Data Assimilation System version 2 (NLDAS2), determine hourly rainfall estimates via the temporal downscaling (disaggregation) of far-more-plentiful daily rain gauge reports. The National Weather Service’s (NWS) National Water Center (NWC) compared NLDAS2 hourly precipitation estimates in the period since 1996, when Next Generation Weather Radar (NEXRAD) Stage-II hourly data became available for use in achieving the disaggregation, with analogous estimates from before 1996, which were disaggregated without the use of radar data. For 20 independently selected days/cases with substantial precipitation during each of the two periods, the post-1996 NLDAS2 hourly amounts were found to be far more highly correlated with cooperative (COOP) rain gauge reports, used for verification, than were the pre-1996 ones. The possibility is then explored that an earlier radar dataset – manually digitized radar (MDR) reports collected from the individual Weather Service Radar-1957 (WSR-57) and WSR-74 radars during the period 1978–1994 – could provide similar benefit in the disaggregation process during the pre-NEXRAD era. In comparisons against COOP hourly reports for 37 warm-season days in the southeast United States and 12 cool-season days in California, it is found that the MDR-disaggregated estimates provided significant statistical improvement over the original, NLDAS2 hourly estimates. Finally, Conterminous United States (CONUS)-wide maps were prepared for each of the four seasons, providing the ratio of hours with measurable precipitation by the MDR methodology against colocated counts of raining hours from surface observation reports. These maps can be used as masks to determine where MDR data would be effective for use in generating any future, hourly precipitation analyses-of-record (AOR), during the 1994-and-prior period.

The paper draws from recent experience with operational regional flash flood warning systems deployed worldwide that utilize radar precipitation data as input to provide regional warnings at high spatial resolution. This paper focuses on radar precipitation data quality control and adjustments in context of making the radar data useful in hydrologic application specific to flash flood warnings. Aspects discussed are radar precipitation preprocessing for hydrologic application, identifying regions of inappropriate radar data for hydrologic use, developing bias adjustment strategies using limited rain gauge networks under the radar umbrella, and impacts of radar precipitation uncertainty on soil water estimates that impact in turn predictions and warnings. The bias adjustment strategies presented are twofold: (1) adjusting for the long-term mean bias in the radar precipitation data (climatological bias adjustment), and (2) accounting for event-specific, temporal deviations from the climatological bias (through dynamic bias adjustment). The development of radar data masks for regions of inappropriate radar data are based on the spatial distribution of the probability of detection of positive precipitation as compared to that produced by rain gauges under the radar umbrella. Climatological bias adjustment of radar data in regions not masked is accomplished through a probability matching procedure that involves transformed data from the rain gauges under the radar umbrella and from their embedding radar grid boxes. This allows for intensity-dependent climatological bias adjustment. An adaptive Kalman filter is used for dynamic bias adjustment of radar data to account for temporal deviations of the bias from the climatological values. The use of adjusted radar data as forcing for the land-surface component of the flash flood warning system produces more realistic soil water state for the flash flood warning systems than the unadjusted radar data does. The discussion uses examples from Romania and Mexico.

In recent years, there have been an increasing number of watersheds that relied on weather radars to provide accurate precipitation information for flood warning purposes. The major impetus for using a radar-based flood warning system is the increased flood warning lead time because of its wider spatial and temporal coverage when compared with rain and stream gauges alone. This paper presents a framework for a radar-based flood alert system (FAS) for the Oyster Creek Watershed to aid the City of Sugar Land in flood forecasting. Precipitation from two actual storm events in 2009 and 2010 were processed in the hydrologic model Hydrologic Engineering Center–hydrologic modeling system (HEC-HMS) to obtain storm hydrographs. After model calibration, the simulated peak flows were then inputted into the hydraulic model, Hydrologic Engineering Center river analysis system (HEC-RAS), to produce water surface elevations. The study includes two flood warning features of Sugar Land’s FAS: the flood plain map library (FPML) and the flood warning indicator (FWI). The FPML has the capability of portraying inundation levels in real time by displaying predelineated floodplain maps that best represent the current condition during storm events, whereas FWI could serve as a first line of defense by delineating specific areas with potential flooding problems within a watershed by using calibrated radar rainfall data. Both features aim to help city officials and emergency personnel plan for flood monitoring, evacuation planning, and traffic routing.

During the period of April 19 to May 4, 2011, an unusually large amount of precipitation fell across the middle to lower portion of the Mississippi River basin and resulted in record flooding in many locations. The greatest amount of moisture, with rainfall totals near or in excess of 508 mm (20 in.), was located throughout northern Arkansas, southern Missouri, far southern Illinois, and western Kentucky. This study determined, using the procedures outlined in the Manual on Estimation of Probable Maximum Precipitation of the World Meteorological Organization (WMO), the extent of how much more precipitation could have occurred in the same region from the same storms if the maximum possible amount of moisture had been available. A detailed examination of the synoptic weather conditions from mid-April through early May in 2011 indicated that two distinct systems affected the region with heavy rainfall during that period. National Weather Service radar and surface observation data from each storm was combined and maximized in accordance with the WMO procedures and compared with the precipitation depths used by the U.S. Army Corps of Engineers for the currently accepted project design flood (PDF). The calculations performed by this in-place maximization study produced values, calculating that the first system could have been 22% greater than observed, whereas an increase of 34% was possible for the second system. The result of maximizing both storms that occurred in the April 19 to May 4, 2011 time frame across the mid-Mississippi River Valley were shown to not exceed the rainfall used for the current Mississippi River and Tributaries (MR-T) Project PDF and demonstrates that the current PDF does not need to be changed on the basis of the observations of this event alone.

An Operational Hydrometeorological Forecast System for Espírito Santo State (HESS), Brazil, is described with special emphasis on civil protection, agriculture and harbor applications, which are heavily dependent on present and future weather conditions between 30 min to 15 days time interval. There is an urgent need for weather services based on high-performance processing systems; ultra high resolution numerical weather prediction systems; higher spatial and temporal resolution monitoring systems; integrated environmental information systems; faster communications systems; and advanced science and technology to yield specific diagnostic and prognostic weather products. The main components of the HESS are described.

Knowledge of how the frequency and intensity of extreme events are affected by climate variability will contribute to answering questions related to drought and water availability that are important in water resources management. A tool titled the “International Center for Integrated Water Resources Management–Regional Analysis of Frequency Tool” (ICI-RAFT), which includes a database of several climate indices, was developed to accomplish this using rainfall observations on the ground. The method used was applied to rainfall obtained from 20 rain gauges in western central Argentina and hurricanes in the Atlantic Ocean during the respective active seasons. The global climate index that was found to have the greatest impact in both cases was the Madden-Julian Oscillation (MJO), which is the focus of the current study. Specifically, MJO activity over central Africa (MJO9) using a lag of 4 months and the western Indian Ocean (MJO10) using zero lag had the greatest influences, respectively. It was found that knowledge of the value of the MJO at these locations allows the prediction of below- or above-average rainfall or hurricane conditions for the current season and even allows a more precise quantitative estimate of this activity to be made.

Urban floods can occur immediately after heavy rainfall depending on the characteristics of the urban environment. Therefore, earlier warnings, even by 5–10 min, are crucial in terms of human safety. In this case study, an X-band multiparameter (X-MP) radar network, which has a 1-min temporal and 250-m spatial resolution, was employed to forecast urban floods in Kobe, Japan. The X-MP radar data can be used for flood prediction when coupled with urban flood alert criteria nomographs. Thus, a specific objective of this study was to develop a flood alert criteria nomograph for the Toga River in Kobe, Japan. To evaluate its applicability, flood events were reproduced using the nomograph and the data were compared with observed rain gauge and radar rainfall data of actual flooding events. Additionally, flood occurrences were predicted using the nomograph and forecasted X-MP radar rainfall data using a short-term rainfall prediction method. It was demonstrated that the method is accurate and can provide adequate lead time for flood warnings.

In this study, a hydrologic web-mapping application for North Carolina State University’s Hofmann Forest is developed using geospatial information science (GIS) resources, measured point precipitation, and radar precipitation data obtained from the National Weather Service (NWS). The development of such a web application will enhance the visualization and manipulation techniques for hydrologic modeling. This paper’s main focus is on the various steps involved in the development of the web-mapping application and the hydrologic analysis. The two different interpolation methods, one a deterministic method, inverse distance weighting (IDW), and another, a probabilistic method, Kriging, are used to obtain NWS precipitation estimates at 14 forest rain gauge locations. The eventual goal of this project is to use the rain gauge data from the Hofmann Forest to calibrate the spatial pattern of daily radar estimates for use in research and forest operations; however, the accuracy of radar estimates has yet to reach that stage of capability.

Spatial precipitation is a major input to distributed hydrological models, and the accuracy of runoff predictions greatly depends on its accuracy. Satellite-based precipitation products are expected to offer an alternative to ground-based rainfall estimates in the present and the foreseeable future. In the present study, the suitability of tropical rainfall measuring mission (TRMM) multisatellite precipitation analysis (TMPA) rainfall in driving a distributed hydrological model for runoff prediction was evaluated. For this purpose, a hydrological model from the literature was calibrated and validated using raingauge data on daily time step for simulation of runoff in Kopili River basin ($=\mathrm{7,198}{\mathrm{km}}^2$) during 2003–2010. The calibrated model was then used for simulation of runoff employing specialized software and compared with the observed discharge at Kherunighat gauging site. Evaluation criteria, i.e., coefficient of correlation (CC), Nash-Sutcliffe coefficient (NSE), percent bias (PBIAS) and root mean square error (RMSE)-observations standard deviation ratio (RSR) were adopted to judge the performance of the model under different rainfall datasets. Simulation using gauge precipitation, the values of CC, NSE, PBIAS, and RSR were found to be 0.85, 0.67, 4.73, and 0.45 respectively during calibration and 092, 0.85, $-1.50$, and 0.29, respectively, during validation indicating overall good model performance. Furthermore, using the raw TMPA precipitation, the values of CC, NSE, PBIAS, and RSR were found to be 0.90, $-5.21$, 141.55, and 2.01, respectively during the period 2004 to 2007 and 0.85, $-13.49$, 179.89, and 2.80, respectively during simulation time period from 2008 to 2010. The moderate value of RSR indicates that the raw TMPA precipitation-based simulation represents the low flow, including rising and recession limbs fairly well, however, in case of the high-flow periods, overpredictions for all years were observed. The evaluation result concluded that the raw TRMM precipitation data are heavily biased from observed precipitation and are incompatible for daily runoff simulation in the study area and bias correction is essential for TRMM precipitation correction. However, after employing adequate bias-correction techniques, the TRMM precipitation performed well with higher degree of accuracy and can be used as an alternative to measured rainfall data due to its high spatial resolution where data are insufficient for runoff prediction, i.e., for ungauged basins. Moreover, performance of the TRMM precipitation improved when simulated in combination with the gauge precipitation. Therefore, along with the efforts to improve satellite-based precipitation-estimation techniques, it is also important to develop more-effective near-real-time precipitation bias-adjustment techniques for hydrological applications.

Quantitative precipitation estimation (QPE) for flash flood forecasting is one of the most commonly used applications of modern meteorological radars. The low-power low-cost small X-band dual-polarization radar system is gaining more interest for urban weather monitoring. This case study presents the installation and application of the first routinely operating dual-polarization X-band radar deployed by the Korea Institute of Civil Engineering and Building Technology (KICT) in South Korea. An overview of this KICT radar’s observations and attenuation correction is presented. The self-consistency criterion is used to examine the dual-polarization data quality. $R(K_{dp})$-based rainfall relations are developed to perform rainfall estimation. The $K_{dp}$ field is robustly obtained using an adaptive estimation methodology. Cross comparisons between rainfall products from the KICT X-band dual-polarization radar, ground rain gauges, and traditional Korea Meteorological Administration (KMA) S-band radar are conducted to demonstrate the performance of this X-band radar for urban rainfall mapping.

An important goal of combining weather radar with rain gauge data is to provide reliable estimates of rainfall rate and accumulation and to further identify intense precipitation and issue flood warnings. Scanning radars provide the ability to observe precipitation over wider areas within shorter timeframes compared to rain gauges, leading to improved situational awareness and more accurate and reliable warnings of future precipitation and flooding events. The focus of this study is on evaluating the performance of the multi-radar multi-sensor (MRMS) system with and without the impact of a local gap filling radar. The challenge of using radar and rain gauges to provide accurate rainfall estimates in complex terrain is investigated. The area of interest is the Russian River basin north of San Francisco, CA, which lies within the National Oceanic and Atmospheric Administration (NOAA) Hydrometeorology Testbed (HMT). In this complex mountainous terrain, the challenge of obtaining reliable quantitative precipitation estimations (QPEs) is hindered by beam blockage and overshooting, as well as the enhancement of rainfall on the windward side of mountain ranges. The effectiveness of several local radars, which include four S-band National Weather Service (NWS) Weather Surveillance Radar–1988 Doppler (WSR-88DP) radars and a C-band gap filling TV station radar (i.e., KPIX), are considered for deriving QPE over this region. The precipitation estimation methodologies used the MRMS algorithms and an independent KPIX-only ($Z-R$) based QPE algorithm. In addition, a time series analysis is conducted in order to illustrate the radar-gauge rainfall difference caused by radar beam height. The sampling relative to precipitation vertical structure is also considered in regards to the depth of the precipitation and the height of the bright band. The quantitative evaluation of different QPE products is presented.

High-resolution quantitative precipitation estimation (QPE) from radar and satellite combined with rain gauges is one of the most important guides for hydrological forecasts. Whereas rain gauges provide accurate measurement at a point, remote sensing helps to retrieve the spatial pattern. An algorithm, named Siprec, has been used to blend rain gauges, radar mosaic data, and satellite Eumetsat/MPE estimates by using Poisson’s equation over two basins in Brazil. The results indicated that Siprec decreased the root mean square error (RMSE) when compared to radar and satellite estimates as well as improved the correlation. Most of the errors were related to precipitation above $10\mathrm{mm}{\mathrm{h}}^{-1}$, due to large spatial variability, typical of deep convection. The solution of Poisson’s equation acts directly on the data received at a certain time, converging the amplitude to the rain gauge values and keeping the spatial distribution of the radar or satellite measurement without a priori adjustments. This is an important advantage in an operational environment because it does not require frequent processing to update the weights like other schemes.